Why so many PhDs?

[ktel]

I've noticed that a majority of people on this forum appear to be entering a PhD program directly from undergrad, and it seems to be the norm in the States.

Up here in Canada I find that somewhat unusual, as most people I know complete a Master's first. There is the option to enter directly from undergrad, and there is also the option to fast track your Master's into a PhD, but it's certainly not the norm.

Anyone have an explanation for me?

[wtncffts]

I've noticed that a majority of people on this forum appear to be entering a PhD program directly from undergrad, and it seems to be the norm in the States.

Up here in Canada I find that somewhat unusual, as most people I know complete a Master's first. There is the option to enter directly from undergrad, and there is also the option to fast track your Master's into a PhD, but it's certainly not the norm.

Anyone have an explanation for me?

Explanation for the norm or for the preponderance of people on this site? The former, I don't know, except to say at least a part of it is probably historical/institutional inertia. The latter, well, most people here are American and/or are applying to US universities.

I would dispute your second point a little. As far as I know, in my field, a Master's is a prerequisite for admission to a PhD program for domestic (Canadian) students. If there is a 'direct entry' option, it's for exceptional, outstanding cases, usually requiring an honours degree. I'm pretty sure the MA requirement is similar for many, if not most, of the social sciences and humanities.

[Ninevah]

I've noticed that a majority of people on this forum appear to be entering a PhD program directly from undergrad, and it seems to be the norm in the States.

Up here in Canada I find that somewhat unusual, as most people I know complete a Master's first. There is the option to enter directly from undergrad, and there is also the option to fast track your Master's into a PhD, but it's certainly not the norm.

Anyone have an explanation for me?

Yeah, that's weird for me too. You get a much better chance to get into Master's than PhD (often 2-4 times better, actually!). Additionally, many, if not most, schools prefer applicants with a Master's degree because these students have done some grad school. The only argument in favor of going straight to the PhD is because Doctoral students get better funding. However, if you apply to departments without PhD program all the money go to the Master's.

[wtncffts]

Additionally, many, if not most, schools prefer applicants with a Master's degree because these students have done some grad school.

I don't think that's true, at least going by the rarity of schools which actually offer stand-alone MA programs. I don't believe any of the eight PhD programs I applied to in the US had an MA program, and I'm certain that most of the other top 50 schools do not.

[Ninevah]

I don't think that's true, at least going by the rarity of schools which actually offer stand-alone MA programs. I don't believe any of the eight PhD programs I applied to in the US had an MA program, and I'm certain that most of the other top 50 schools do not.

It's probably field-specific, but many of the universities say something like "In most cases, a completed master's degree is required for admission to the program" (Penn State), "It is suggested but not required that students have or will have a master's degree" (Urbana), etc.

[robot_hamster]

It really depends on your area of study. There are some programs that are only offered as straight through PhD's, while others are Masters then PhD.

[LJK]

I totally agree it depends on your area of study. In psychology there is a debate about whether a MS is actually helpful. For some, it shows that the student is serious, knows what they are getting into, and knows what they want to do. For others, it says that they student has already started to form ways of doing things that will need to be broken down and retrained before anything useful can come out of them. I think part of the reason for the second position is because a lot of MA/MS degrees come from 'lower tier' universities that often don't have phd programs. Getting the MA/MS on the way to the PhD gives the program and the advisor more control over the students training. In fact, many programs don't care that you have already taken a significant number of grad courses - they want you to take their grad courses and will only let you transfer in a few credits. So the trust in the quality of the training issue could account for why many top programs don't offer just PhDs after coming in with a MA/MS. As to why students want to go straight into the PhD? That means you will be done in 4-6 years (or the norm for your field) rather than 1-2 + 4-6 years. Of all the PhD students I know who got MS degrees before starting the PhD, none are taking less than the standard 5 years in my field to get the PhD, even though they did 2 year programs before hand. Who wants to be paid on a graduate student stipend forever?? (I am finishing up my MS and heading on to a PhD now, so I guess I do )

[Kitkat]

With a lot of the hard sciences here I think there are several reasons people go for the PhD. One is because with a lot of jobs you'll need the PhD to get the jobs you want so you might as well get it. Also, many of the PhD programs you get your masters along the way, its just the way that they work. And third, in the sciences at least, schools give you funding if you are in a PhD program, but not if you are going into just a masters.

[hejduk]

I've noticed that a majority of people on this forum appear to be entering a PhD program directly from undergrad, and it seems to be the norm in the States.

Up here in Canada I find that somewhat unusual, as most people I know complete a Master's first. There is the option to enter directly from undergrad, and there is also the option to fast track your Master's into a PhD, but it's certainly not the norm.

Anyone have an explanation for me?

I think this may be dependent on the field, as I'm in communication, and most (at least those I know) go into an MA, and then onto doctoral studies. It's how I'm doing it, and how all my friends have as well.

[StrangeLight]

as a canadian that went to the US to do a combined MA/PhD program...

in canada, schools will give you funding or TA positions to pay for your MA. in the US, terminal MAs are rarely, if ever, funded, so it's better to go to a combined MA/PhD program than to pay out of pocket. many students also use canadian MAs to get positions at US PhD programs (at least in my field). it's a good springboard if a US PhD offer doesn't come up during the application process. but, short answer? you can get paid to do an MA in canada, but you (usually) have to pay to do an MA in the US.

[Ninevah]

as a canadian that went to the US to do a combined MA/PhD program...

in canada, schools will give you funding or TA positions to pay for your MA. in the US, terminal MAs are rarely, if ever, funded, so it's better to go to a combined MA/PhD program than to pay out of pocket. many students also use canadian MAs to get positions at US PhD programs (at least in my field). it's a good springboard if a US PhD offer doesn't come up during the application process. but, short answer? you can get paid to do an MA in canada, but you (usually) have to pay to do an MA in the US.

My understanding is that Master's are only terminal in professional programs (e.g., MBA). Since the school expect you to make a truckload of money with the professional degree, they don't give you money. MA in History, for example, would not be a terminal degree because you are more or less expected to do a PhD after that. You can get funded in an MA program in the US - I was and I know many people in Comm and Poli-Sci that are.

[newms]

I agree with those here who have said that it's very field specific. In my field and in a lot of science fields, if you ultimately want to get a PhD, universities in the US encourage you to apply directly to the PhD rather than applying to the Master's. In my field, many of the top universities do not offer research based Master's and only offer terminal Master's for those interested in working in industry. Also, in my field and in many (all?) of the science fields, schools in the US fund PhD students (wholly or partially) while funding is rare for Master's students. It's different in Canada though (at least in my field), where Master's students are funded and schools encourage you to get a MS before applying for a PhD.

Edited June 29, 2011 by newms

[ktel]

Thanks for all the insight, I find the information interesting.

I have yet to decide whether I will do a PhD, so I am obviously happy to be enrolled in an MASc with funding. I'm also not sure my application would have been strong enough for direct entry into a PhD program, as I really only decided I wanted to do grad school in my final year of undergrad. As a result I took on all my research projects during my last year, after applications would've been sent in.

[starmaker]

In engineering and computer science, I think it is more common to have done a master's first (and possibly to be coming back to academia from industry) than it is in the other sciences and the humanities. Some programs even have separate post-bachelor's and post-master's tracks (the post-master's track requires less coursework).

I will have done a master's first, but that's because I had low undergrad grades, switched fields after undergrad, and have been working in industry while pursuing my master's part-time. The MS gave me a chance to do better in classes and build up more background in my post-undergrad field, plus doing it part-time while working let me save money.

[natsteel]

My understanding is that Master's are only terminal in professional programs (e.g., MBA). Since the school expect you to make a truckload of money with the professional degree, they don't give you money. MA in History, for example, would not be a terminal degree because you are more or less expected to do a PhD after that. You can get funded in an MA program in the US - I was and I know many people in Comm and Poli-Sci that are.

It's misleading to overgeneralize, at least in regards to History. Nevertheless, some schools, even ones that have full PhD programs, offer terminal MAs in History, but those programs are often either looked at as cash cows (i.e., no funding) or largely as an afterthought. Then there are tons of regional colleges and universities which do not have PhD programs, which offer terminal MAs in History.

In History it seems that some (though of course not all) students do an MA to improve on a less-than-stellar undergraduate record or if they are unsure if graduate school is the right decision. I don't think having an MA makes a big difference when applying to top 50 programs in History. Or I should say that it doesn't appear that MA-holders are considered to have a distinct advantage in the admissions process. But these are just my own perceptions...

Edited June 30, 2011 by natsteel

[Behavioral]

It's misleading to overgeneralize, at least in regards to History. Nevertheless, some schools, even ones that have full PhD programs, offer terminal MAs in History, but those programs are often either looked at as cash cows (i.e., no funding) or largely as an afterthought. Then there are tons of regional colleges and universities which do not have PhD programs, which offer terminal MAs in History.

In History it seems that some (though of course not all) students do an MA to improve on a less-than-stellar undergraduate record or if they are unsure if graduate school is the right decision. I don't think having an MA makes a big difference when applying to top 50 programs in History. Or I should say that it doesn't appear that MA-holders are considered to have a distinct advantage in the admissions process. But these are just my own perceptions...

I would take a more Spence-ian signaling approach to that--I think that the majority of people who do tend to get MAs in history are the ones who are trying to rectify a "poor" undergraduate record--if you use undergraduate success as a predictor for graduate success (I feel there's a significant correlation here, especially since undergraduate coursework and overall responsibilities are more lax), then you see a lot of regression to mediocrity before applying to Ph.D. programs. IF Ph.D. programs have noticed this, then they probably scrutinize MA-holders' applications a bit more.

Of course this is just a hypothesis, but it'd make sense since it would at least partially explain why many with MAs don't get into better-ranked programs (regardless of field) even though we may posit that an MA provides superior "training" than just a Bachelors.

[Ninevah]

I would take a more Spence-ian signaling approach to that--I think that the majority of people who do tend to get MAs in history are the ones who are trying to rectify a "poor" undergraduate record--if you use undergraduate success as a predictor for graduate success (I feel there's a significant correlation here, especially since undergraduate coursework and overall responsibilities are more lax), then you see a lot of regression to mediocrity before applying to Ph.D. programs. IF Ph.D. programs have noticed this, then they probably scrutinize MA-holders' applications a bit more.

Of course this is just a hypothesis, but it'd make sense since it would at least partially explain why many with MAs don't get into better-ranked programs (regardless of field) even though we may posit that an MA provides superior "training" than just a Bachelors.

Sure, that might be a reason why people go for MA after PhD. But while undergrad GPA might be a reliable predictor for success (sure, if you paid more attention and did more work in class you are very likely to be better a student), isn't it true that graduate classes are harder? If you can deal with harder classes better than an undergrad with his BA classes, isn't that an even more reliable predictor? Of course, your graduate GPA will be more important than your UGPA.

Your UGPA might be lower than someone who went straight to PhD, but we (and I assume the adcomms) know how your first couple of years in BA with all their temptation can distract you from academics. By the time you finish the MA you'll be 4-5 years away from that point. How much weight would you put to those initial grades?

As for your last paragraph, I don't know where you get that info. it would be nice to see some statistic here, though I doubt anything is available. Of course many MAs don't get into top programs. Many BAs don't too. Nothing is the "silver bullet" for success.

Edited June 30, 2011 by Ninevah

[starmaker]

Sure, that might be a reason why people go for MA after PhD. But while undergrad GPA might be a reliable predictor for success (sure, if you paid more attention and did more work in class you are very likely to be better a student), isn't it true that graduate classes are harder? If you can deal with harder classes better than an undergrad with his BA classes, isn't that an even more reliable predictor? Of course, your graduate GPA will be more important than your UGPA.

My grad classes are not harder than my undergrad classes were (I am doing an MS before a PhD in part because I had a poor undergrad GPA). I attribute this to the schools in question. Also, it's not uncommon for grad grades to be a little inflated.

On the other hand, UGPA is not a great predictor of grad school success. It might be a predictor of grad coursework success, but the qualities needed to get good grades in classes and the qualities needed to be a good researcher are not the same (more power to you if you have both sets).

Though, like I said before, my field is one where getting an MS before a PhD is relatively common. I rather wish adcoms would give extra scrutiny to my app, as the non-undergrad-grades parts are pretty strong.

[ZeeMore21]

I would take a more Spence-ian signaling approach to that--I think that the majority of people who do tend to get MAs in history are the ones who are trying to rectify a "poor" undergraduate record--if you use undergraduate success as a predictor for graduate success (I feel there's a significant correlation here, especially since undergraduate coursework and overall responsibilities are more lax), then you see a lot of regression to mediocrity before applying to Ph.D. programs. IF Ph.D. programs have noticed this, then they probably scrutinize MA-holders' applications a bit more.

Of course this is just a hypothesis, but it'd make sense since it would at least partially explain why many with MAs don't get into better-ranked programs (regardless of field) even though we may posit that an MA provides superior "training" than just a Bachelors.

I think your claim that MAs don't get into top programs is bogus, and I take offense to your generalization being someone who did get an M.A before starting a PhD program. Plus, since you're in marketing and are NOT in the History field, I doubt you have any right to make such careless assumptions.

Please don't belittle the hard work that goes into getting an M.A. And trust me, M.As aren't losers like you claim, or else they wouldn't have even gotten into graduate school. You might not see, but I'm sure admissions committees do see promise in people who might have struggled in their undergrad. I've had a mentor of mine get a 2.3 GPA and end up completing a PhD at Penn, and is now the head of a department. So your argument that those who get M.As because they had mediocre undergrad scores don't get into top programs is completely false. Your opinion that all those who go to pursue an M.A must have been mediocre or will not succeed is also false.

Sorry if this is harsh, just tired of people feeling the need to put others down. Seems as though you are not considering the fact that people with M.As might come across your post and be offended by your generalizations. I've gotten into a school that is at the top of my field after completing my M.A so I think your argument does not hold.

Edited July 1, 2011 by ZeeMore21

[hejduk]

I would take a more Spence-ian signaling approach to that--I think that the majority of people who do tend to get MAs in history are the ones who are trying to rectify a "poor" undergraduate record--if you use undergraduate success as a predictor for graduate success (I feel there's a significant correlation here, especially since undergraduate coursework and overall responsibilities are more lax), then you see a lot of regression to mediocrity before applying to Ph.D. programs. IF Ph.D. programs have noticed this, then they probably scrutinize MA-holders' applications a bit more.

Of course this is just a hypothesis, but it'd make sense since it would at least partially explain why many with MAs don't get into better-ranked programs (regardless of field) even though we may posit that an MA provides superior "training" than just a Bachelors.

IMHO, I'd have to agree that the rationale here is flawed. While I may not have attended an "Ivy League" (nor did I have a desire to), my pursuit of an MA was the practical next step after undergrad. By pursuing my MA, I wasn't trying to hide a blemished undergrad record or trying to make myself look better for a doctoral program.

I know people in my field (Communication) that go from BA to PhD, or BA to MA to PhD, so one path doesn't fit all. I think there are too many factors involved in deciding which path to take, so it's probably tough and unfair to generalize specific discipline or field.

[ZeeMore21]

I think your claim that MAs don't get into top programs is bogus, and I take offense to your generalization being someone who did get an M.A before starting a PhD program. Plus, since you're in marketing and are NOT in the History field, I doubt you have any right to make such careless assumptions.

Please don't belittle the hard work that goes into getting an M.A. And trust me, M.As aren't losers like you claim, or else they wouldn't have even gotten into graduate school. You might not see, but I'm sure admissions committees do see promise in people who might have struggled in their undergrad. I've had a mentor of mine get a 2.3 GPA and end up completing a PhD at Penn, and is now the head of a department. So your argument that those who get M.As because they had mediocre undergrad scores don't get into top programs is completely false. Your opinion that all those who go to pursue an M.A must have been mediocre or will not succeed is also false.

Sorry if this is harsh, just tired of people feeling the need to put others down. Seems as though you are not considering the fact that people with M.As might come across your post and be offended by your generalizations. I've gotten into a school that is at the top of my field after completing my M.A so I think your argument does not hold.

Would like to know what the person who voted me down's logic is. I think it is unnecessary.

[Behavioral]

Sure, that might be a reason why people go for MA after PhD. But while undergrad GPA might be a reliable predictor for success (sure, if you paid more attention and did more work in class you are very likely to be better a student), isn't it true that graduate classes are harder? If you can deal with harder classes better than an undergrad with his BA classes, isn't that an even more reliable predictor? Of course, your graduate GPA will be more important than your UGPA.

Your UGPA might be lower than someone who went straight to PhD, but we (and I assume the adcomms) know how your first couple of years in BA with all their temptation can distract you from academics. By the time you finish the MA you'll be 4-5 years away from that point. How much weight would you put to those initial grades?

As for your last paragraph, I don't know where you get that info. it would be nice to see some statistic here, though I doubt anything is available. Of course many MAs don't get into top programs. Many BAs don't too. Nothing is the "silver bullet" for success.

This is the first of my apologies for my post -- if I offended anyone, I truly resign my post. As someone with training in microeconomic theory (namely in bounded rationality and incomplete information modeling), this is pretty much what Michael Spence's signalling model would (at least in my relatively obtuse understanding of it) predict, given that if available information (i.e., GPA, research experience, letters of rec, etc.) is used as a diagnostic for success/attractiveness as a candidate (grad student, employee, etc.), then exogenously increasing desirable (that is, holding all other variables equal) traits emits a stronger positive signal to the agent making the hiring/acceptance decision. That was a mouthful--sorry. In this (theoretical) model, people are discretely distributed into two groups (typically in economics "high efficient" and "low efficient" people), which in this case would be translated to "people who would succeed in our program if admitted" and "people who would NOT succeed in our program if admitted." If this were the case, there are various 'signal-expense' schedules each group can take to meet an equilibrium [(effort exerted to emit signal) = (probability of success)(utility gained from admission)] => if you increase a person's perceived probability that they will succeed (i.e., not fail) at something, they will alter their effort accordingly; this finding is what netted Michael Spence a Nobel Prize in economics since it provided an elegant and intuitive model as to why employers prefer college graduates over non-college graduates for jobs that don't utilize the skills/knowledge gained from a college education (in other words, how college education can be perceived as a 'signal' rather than actual job preparation/training). In this context, a graduate school would look at undergraduate success (even if undergraduate GPA is NOT a strong predictor for graduate school success) as a signal for admission. Controversial? Yes. Spence's signalling model is still contested by economists, psychologists, liberals, conservatives, etc. as to whether it's actually the correct normative model people employ; however, empirically speaking, it's plausible--at least enough for him to win the Nobel Prize.

http://nobelprize.org/nobel_prizes/economics/laureates/2001/spence-lecture.pdf

There's also the positive vs. negative framing literature that could be at play whether positive attributes or negative attributes in candidates are weighted more heavily in admissions.

I think your claim that MAs don't get into top programs is bogus, and I take offense to your generalization being someone who did get an M.A before starting a PhD program. Plus, since you're in marketing and are NOT in the History field, I doubt you have any right to make such careless assumptions.

Please don't belittle the hard work that goes into getting an M.A. And trust me, M.As aren't losers like you claim, or else they wouldn't have even gotten into graduate school. You might not see, but I'm sure admissions committees do see promise in people who might have struggled in their undergrad. I've had a mentor of mine get a 2.3 GPA and end up completing a PhD at Penn, and is now the head of a department. So your argument that those who get M.As because they had mediocre undergrad scores don't get into top programs is completely false. Your opinion that all those who go to pursue an M.A must have been mediocre or will not succeed is also false.

Sorry if this is harsh, just tired of people feeling the need to put others down. Seems as though you are not considering the fact that people with M.As might come across your post and be offended by your generalizations. I've gotten into a school that is at the top of my field after completing my M.A so I think your argument does not hold.

Once again, I do apologize for any offensive statements--especially when done outside my field. I do come from a background (Joint Math/Econ and Psychology) where people going into both Economics and Psychology Ph.D. programs typically do so with just a baccalaureate degree--in both fields, getting a terminal masters (if one knows they're going for a Ph.D. anyway) is a waste of money unless it's being leveraged to help one draw emphasis on graduate success over a less-than-perfect undergraduate career.

Now, you're also twisting my words and turning them into absolutes. I didn't claim that someone with an M.A. couldn't become successful--I didn't even claim that someone with an M.A. couldn't get into a top program. I'm careful not to make mistakes based on absolutes since they're easy to refute using just one counterexample (you provided one, and I count conjure up several from my undergraduate institution). Secondly, I never once claimed that the applicants themselves (as personal characteristics) were mediocre; I referred to their applications. I definitely don't make the claim that exceptional people necessarily have to be good at school or the converse that those who are good at school are exceptional people. I think you threw that argument in my mouth.

IMHO, I'd have to agree that the rationale here is flawed. While I may not have attended an "Ivy League" (nor did I have a desire to), my pursuit of an MA was the practical next step after undergrad. By pursuing my MA, I wasn't trying to hide a blemished undergrad record or trying to make myself look better for a doctoral program.

I know people in my field (Communication) that go from BA to PhD, or BA to MA to PhD, so one path doesn't fit all. I think there are too many factors involved in deciding which path to take, so it's probably tough and unfair to generalize specific discipline or field.

Once again, very sorry for any offhanded claims and remarks. I based them on my fields (which I didn't cite as qualifiers), so this doesn't necessarily apply to other fields/situations. I know that the choice between BA -> MA -> PhD and BA -> PhD is complex, but another mistake of mine is that I didn't claim the assumption that those who would eventually apply to PhD programs were sure that they were going to pursue a doctorate for a large portion of their undergraduate degree (leaving enough time for the last 2-3 years to rectify a poor early start; as in my case when I started with a 2.7 my first 4 quarters, and finished my degree with honors and a 3.7). A lot of my friends with very competitive undergraduate records do choose to go to Masters programs to 'test the waters', either seeing whether or not they like research; if they want to work in industry for a while before possibly going into academia and want a leg up on more technical or senior positions; or what have you.

---

If I offended anyone else, I'm very sorry. I had no intentions of belittling any persons or credentials -- hell, I don't even have a Masters of my own yet. I was merely trying to apply a widely used economic theory into the given context provided by this thread, and it turned quite hairy.

[Behavioral]

I rather wish adcoms would give extra scrutiny to my app, as the non-undergrad-grades parts are pretty strong.

I'm all for this, too, but it seems to be not the case for certain disciplines. The system that's used in Canada and Europe (maybe other places, too) where a Masters pre-doctoral application is the norm makes a lot more sense. Undergraduate studies could hardly be any MORE different than graduate school, both in terms of difficulty in classes, the different type of thinking/analysis involved in many graduate programs, and the responsibilities involved (research sometimes not being a necessary [or at least a big] component in undergraduate studies). Having that buffer in the middle as preparation, in my opinion, would help the problem of attrition at many schools. Students that came straight out of undergrad that drop out because they couldn't handle the work, didn't know if research was for them, etc., may have avoided it while getting their Masters and then choose to not pursue any more education; schools wouldn't be investing resources (finances, professor time and effort training students, etc.); open spots for doctoral programs would mainly be open to those who've already gotten a taste of graduate school and are at least a bit more informed as to what a doctoral program entails vs. someone with no experience in graduate school. Taken together, I don't know why schools in the US take as many baccalaureates as they do (which goes back to the original question in hand).

More strikingly, some fields (like law), graduate degrees play essentially no role in admissions (LSAT and undergrad GPA pretty much dictate the admission game)--perhaps because there's a bit of politicking between schools since average undergrad GPA (and not graduate GPA) is a determinant for USNEWS ranking.

[ZeeMore21]

I'm not familiar with the Spencer model that your are proposing Behavioral, but I doubt that school committees go by that model when they make their decisions. There isn't a cut and dry formula they use, which explains why there are many applicants with 4.0 GPAs and perfect GREs that still don't get admitted. Having passion in your field of study goes a long way, and it doesn't necessarily have to be conveyed at the quantitative level. Materials such as a writing sample and personal statement are weighed much more heavily in the humanities. If you have great grades but can't write or communicate effectively, you don't belong in a doctorate program.

[Behavioral]

I'm not familiar with the Spencer model that your are proposing Behavioral, but I doubt that school committees go by that model when they make their decisions. There isn't a cut and dry formula they use, which explains why there are many applicants with 4.0 GPAs and perfect GREs that still don't get admitted. Having passion in your field of study goes a long way, and it doesn't necessarily have to be conveyed at the quantitative level. Materials such as a writing sample and personal statement are weighed much more heavily in the humanities. If you have great grades but can't write or communicate effectively, you don't belong in a doctorate program.

That's why there are very few regression models that only use one explanatory variable. The signalling model doesn't use just one predictor, but weights a number of predictors accordingly (i.e., quantity and however you would like to "quantify" quality of research experience is likely to trump GPA for Ph.D. admissions). Also, there is noise within signalling models, which is why Signal Detection Theory was proposed; essentially it can become a Bayesian statistics problem as you increase the amount of trials, gaining information between variation in the predictors (i.e., the ranges of accepted individuals with various stats/grades/etc.), and their success in the program => if you notice that students with a balanced (fairly high) GPA along with a thorough research background seem to do well, but those with high GPA+no research nor high research+very low GPA don't do well, then you'll begin incorporating this in your model to find 'ideal' candidates; if there's a way to quantify SOP quality, too, for example, by assigning a grade, a binomial (good vs. bad) descriptor, or by using a Likert scale, this can, too, be translated into a regression model.